#!/usr/bin/perl -w
# Copyright © 2006 Jeremy English <jhe@jeremyenglish.org>
# Script to pull images out of flickr's rss 2.0 files.
#
# Permission to use, copy, modify, distribute, and sell this software and its
# documentation for any purpose is hereby granted without fee, provided that
# the above copyright notice appear in all copies and that both that
# copyright notice and this permission notice appear in supporting
# documentation.  No representations are made about the suitability of this
# software for any purpose.  It is provided "as is" without express or 
# implied warranty.
# 
# Created: 12-January-2006 



use strict;

my $progname =$0; $progname =~ s@.*/@@g;

my $verbose = 0;

#Pull feeds listed
#Pull the image source tags
#If the image dosen't exist in the destination folder then download it

sub pull_pixs {
  my ($url) = @_;
  my $expirey = 30; #only download every 30 minutes;

  my $now = time;

  $_ = $url;
  s@/+$@@;
  s@^http://@@gi;
  s@[^-_a-z\d]@_@gi;
  my $file = $_;
#  my $hfile = "$html_cache_dir/$_";
  my $hfile = "$_";

  # lifted from cheesegrater.pl @ www.jwz.org
  # check the expirey, and if we've checked this URL recently, don't
  # re-download the content.
  #
  {
    my ($dev,$ino,$mode,$nlink,$uid,$gid,$rdev,$size,
        $atime,$mtime,$ctime,$blksize,$blocks) = stat($hfile);
    if (!defined ($mtime)) {
      print STDERR "$progname: $hfile does not exist\n" if ($verbose > 3);
    } elsif ($size < 512) {
      print STDERR "$progname: $hfile is only $size bytes\n" if ($verbose > 3);
    } else {
      my $minutes_ago = int (($now - $mtime) / 60);
      if ($minutes_ago < $expirey) {
        print STDERR "$progname: $hfile: modified < $expirey ($minutes_ago) " .
          "minutes ago.\n"
            if ($verbose > 2);
        return $file;
      } else {
        print STDERR "$progname: $hfile last modified $minutes_ago " .
          "minutes ago.\n"
            if ($verbose > 3);
      }
    }
  }

  my @cmd = ("wget", "-O", $hfile, $url);

  if (system(@cmd) == 0) {
    #pull images
    open(IN, "< $hfile");
    while(<IN>){
      my $image = "";
      if (/ *< *media:content *url *= *"([^"]+).*/){
	$image = $1;
	if ($image ne ""){
	  my $local_name = $image;
	  $local_name =~ s@.*/@@g;
	  if (! (-e $local_name )){
	    my @cmd = ("wget", $image);
	    system(@cmd);
	  }
	}
      }
    }
    close(IN);
  }
  return $file;
}

sub usage {
  print STDERR "usage: $progname [urls...]\n";
  exit 1;
}

sub main {
  my @urls = ();
  while ($_ = $ARGV[0]) {
    shift @ARGV;
    if ($_ eq "--verbose"){ $verbose++; }
    elsif (m/^-v+$/) { $verbose += length($_)-1; }
    elsif (m/^-./){ usage; }
    elsif (m/^http:/) { push @urls, $_; }
    else { usage; }
  }

  usage unless ($#urls >= 0);

  foreach (@urls){
    pull_pixs($_);
  }

  exit 0;
}

main;
exit 0;